Answers are grounded in this paper and cite their sources.
Synapse-resolved connectomes of Drosophila melanogaster now map the entire adult brain and ventral nerve cord, but behavior also depends on synaptic strengths, neural dynamics, and the body. A connectome is therefore better used as a testable structural prior than as a blueprint to copy. This review examines how digital Drosophila (connectomes plus the models and body simulators built on them) can serve this role in embodied control. We trace how electron-microscopy images become a connectivity graph, where errors enter, and how heterogeneous datasets can be combined reproducibly. We then compare three uses of the graph (extracting a circuit principle, embedding a reconstructed module, or adopting the whole connectome as controller topology) with one another and with conventional controllers, separating experimentally validated findings from computational predictions. Connectome-constrained models can predict neural activity and sensorimotor pathways, but simulated success does not by itself establish biological accuracy, and sensory encoding, motor decoding, and body mechanics are as much part of the controller as the wiring. From this synthesis we derive five candidate design principles (sparse modular communication, recurrent state estimation, hierarchical motor control, selective plasticity, and body–controller co-design) and one methodological principle (anatomically indexed intervention), each paired with falsifiable predictions, matched baselines, and sim-to-real stress tests.
No takes yet. Share an insight, caveat, or question.
Jiang et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: